1. Field of the Invention
This invention relates in general to the field of interferometry and, in particular, to a novel interferometric methods and interferometers for simultaneously obtaining the advantages of both white light and high coherence interferometry by modulating the spectral distribution of the light source.
2. Description of the Prior Art
High-accuracy shape measurements are an indispensable part of modern technology. In particular, interferometric methods rely on the interference of two beams of light to produce interference patterns from which measurement information about a test surface can be extracted. One beam, normally referred to as the object beam, is reflected from the test surface and carries information about its topography or other property. The other, reference, beam is a standard against which the object beam is compared and is typically produced by reflecting light from a surface of known geometry, such as a flat or a spherical surface (normally referred to in the art as reference flat and reference or transmission sphere, respectively). The two beams are combined to cause them to interfere and the shape of the test part is recovered by analyzing the interference pattern with methods well known in the art, such as phase-shifting or carrier-fringe algorithms.
All interferometric techniques rely on the properties of the light source, such as its spectral and spatial distribution, to form interference fringes from which surface characteristics of the test object are recovered. There are several types of interferometers, typically categorized by the geometrical arrangement of the object and reference arms and by the type of light source, namely the so-called temporal and spatial coherence of the light. Interferometers that use high temporal coherence sources are normally referred to as laser interferometers, while instruments that use temporally incoherent (or low coherence) sources are referred to as white-light interferometers (WLI). One of the most significant distinctions between these two types of interferometer is the localization of the interference fringes. WLIs produce fringes that are only visible in a limited space around the location where the optical path length difference (OPD) between the object and the reference beams is close to zero; i.e., where the delay between the object and reference beams is very small (localized fringes). In laser interferometry the fringes are formed for much larger OPD values; thus, the fringes are not localized but they are repeated periodically as the OPD varies with a period of one wavelength of the source.
A common implementation of white-light interferometry is the Michelson interferometer, illustrated schematically in FIG. 1. The white light L emitted from a source 10 is split by a beam-splitter 12 into an object beam O and a reference beam R. The object beam is directed toward and reflected back to the beam-splitter by the surface 14 of a test object, while the reference beam is directed toward and reflected back by a reference mirror 16 of known shape. Upon passing through the beam-splitter 12, the two beams are recombined to produce interference fringes that reach maximum intensity when the OPD between the reference and object arms of the interferometer (i.e., the difference in their distance from the beam-splitter) is reduced to zero. A scanning mechanism 18 is used to change the position of the test object (or, alternatively, of the reference mirror) so that the OPD at every measured point (or pixel) on the surface of the object can be reduced to substantially zero during a measurement. The resulting interferogram I is detected by a detector 20 and recorded for analysis by a processor 22. Because white-light interferometers produce fringes only around the zero OPD location, they are suitable for so-called scanning interferometry. This is because all portions of the surface of the sample are measured with a single scan as the respective interference fringes, which only occur within the coherence length of the light, are thereby localized at the detector 20.
The Fizeau interferometer configuration is often used in laser and narrowband interferometry. As illustrated schematically in FIG. 2, a monochromatic or narrowband light beam L′ typically produced by a laser source 24 is reflected by a beam-splitter 26 toward a transparent reference flat 28 and an axially aligned object surface 30. Upon reflection of the light L′ from each surface, a reference beam R and an object beam O are produced and returned on axis toward the beam-splitter 28. The beams are recombined, thereby producing interference, and are passed back through the beam-splitter to a detector 32 and processor 34 for recordation and analysis. A shifting mechanism 36 is provided to shift the position of the test object (or the reference mirror) so that phase-shifting measurements can be carried out. Because of the multiple fringes produced by laser interferometers, they are best suited for measuring smooth surfaces without discontinuities.
Each type of interferometer has strengths over the other, but also weaknesses that render it unsuitable for particular applications that are best addressed by the other. For example, Fizeau interferometers are particularly valuable for their common-path configuration that greatly reduces measurement errors caused by the interferometer's optical system. However, they are susceptible to coherence noise, such as speckle and diffraction patterns caused by contamination, because high-coherence light has the ability to interfere even when the OPD is quite large. Examples of such noise are typical concentric fringe patterns stemming from diffraction on dust particles that may be present on the active optical surfaces of the interferometer (so-called bulls eyes). As a result, laser interferometers require extremely clean interiors and assembly in clean rooms. Another common problem with laser interferometry is the stability and the coherence length of the laser source. Often, especially for gas lasers with long cavity, the source emits multiple longitudinal modes that limit the total OPD of coherence. Short cavity gas lasers can produce a single longitudinal mode that provides much longer coherence lengths but that can also exhibit power and coherence length fluctuations.
White-light interferometers are generally free from coherence noise and stability. Because they produce only localized fringes, they can be designed to avoid interference from spurious reflections, such as from the intervening surfaces of glass blocks disposed in the optical path. However, WLIs require careful balancing of the OPD in order to produce fringes at the desired location and their scanning mechanism and moving parts require a more complicated mechanical design to ensure a precise scanning motion. These design requirements are significantly less stringent in laser interferometers.
In view of the foregoing, an interferometric method and an interferometer that provide the advantages of both laser and white-light interferometry would be very desirable and would represent a significant advance in the art.